+ All Categories
Home > Documents > Urban/Rural Areas

Urban/Rural Areas

Date post: 23-Jan-2022
Category:
Upload: others
View: 32 times
Download: 0 times
Share this document with a friend
34
Urban/Rural Areas Population density (from a 1 km 2 grid), land cover and remoteness as basic elements for an urban/rural typology at LAU2 level (In progress…) Francisco J. Goerlich, University of Valencia and Ivie Isidro Cantarino, Polytechnic University of Valencia EFGS Conference - Sofia, October 23-25 th , 2013 1
Transcript
Page 1: Urban/Rural Areas

Urban/Rural AreasPopulation density (from a 1 km2 grid),

land cover and remoteness as basic elements for an urban/rural typology at

LAU2 level

(In progress…)

Francisco J. Goerlich, University of Valencia and IvieIsidro Cantarino, Polytechnic University of Valencia

EFGS Conference - Sofia, October 23-25th, 2013

1

Page 2: Urban/Rural Areas

Approaches to define the urban/rural character of communes… There is no general agreement on how to

classify urban/rural communes… so many classifications rules, typical employed by public bodies and statistical agencies, rely on thresholds in population density, settlement size or a combination of both.

The old OECD classification rule identify rural communes as those with less than 150 inhab/km2, or a typical rule used in Spain is to consider urban communes those as having at least 10.000 inhabitants.

2

Page 3: Urban/Rural Areas

3

OECD criteria: a commune is rural if its population density is lower than 150 inhab./km2.

Rural LAU2: 87.1% (7,066)

Population: 24.9%

Territory: 91.1%

Minimum population size criteria: a commune is urban if its has at least 10.000 inhabitants, and rural otherwise.

Rural LAU2: 91.2% (7,399)Population: 21.9%Territory: 80.9%

Page 4: Urban/Rural Areas

Demography from a 1-km2 population grid…

4

Recently, a great effort has been done from Eurostat, DG-Regio and the OECD in order to define a consistent system

of urban/rural typologies that starts from a 1 km2 population grid.

The key element is to distinguish between urban and rural cells.

Page 5: Urban/Rural Areas

Urban and rural cells… The three key concepts are:

1. Rural cells: Grid cells outside urban clusters. Rural grid cells can be inhabited or not.

2. Urban clusters: Clusters of contiguous, including diagonals, grid cells of 1 km2 with a density of at least 300 inhabitants per km2 and a minimum population of 5,000 inhabitants.

3. Urban centers or High density clusters (city center): Clusters of contiguous, excluding diagonals but filling gaps, grid cells of 1 km2 with a density of at least 1,500 inhabitants per km2 and a minimum population of 50,000 inhabitants.

5

Page 6: Urban/Rural Areas

6

The grid…

Page 7: Urban/Rural Areas

7

Rural population:20.3% (9.05 million)

Urban population (living in Urban Clusters):79.7% (35.65 million)UC: 737

…the rural/urban cells…

Page 8: Urban/Rural Areas

8

Some Urban Clusters…Madrid

Barcelona

Alicante/Murcia

Page 9: Urban/Rural Areas

9

Population living in High Density Clusters (Urban Centers):50.0% (22.35 million)HDC: 105

…and the high density clusters.

LAU2Number % People % Support

Rural Areas 84,449 89.0% 9,054,928 20.3%Urban clusters (UC) 10,467 11.0% 35,654,036 79.7% 737 1,493Total 94,916 100.0% 44,708,964 100.0%

Urban Centers (HDC) 2,463 2.6% 22,348,890 50.0% 105 287Source: Own elaboration.

Table 1. Population distribution acording to the types of cells amd number of clusters.Cells Population Clusters

Page 10: Urban/Rural Areas

10

MadridSome High Density Clusters…

Valencia

Page 11: Urban/Rural Areas

From urban/rural cells to urban/rural communes… The rules are based on the share of population living in

rural cells and the different type of clusters

1. Rural commune (area) or thinly populated area, if at least 50% of the population lives in rural grid cells.

2. Small urban commune (area), towns and suburbs or intermediate density area, if less than 50% of the population lives in rural grid cells and less than 50% of the population lives in high density clusters.

3. Large urban commune (area), cities or densely populated area, if at least 50% of the population lives in high density clusters.

11

Page 12: Urban/Rural Areas

Number % Total Total (%) Rural  Rural (%) Urban  Urban (%) Km² Km² (%)Urban 220 2.7% 24,002,578 53.7% 554,491 6.1% 23,448,087 65.8% 22,975 4.6%Intermediate 1,025 12.6% 13,635,414 30.5% 1,628,659 18.0% 12,006,755 33.7% 100,068 19.8%Rural 6,865 84.6% 7,070,972 15.8% 6,871,778 75.9% 199,194 0.6% 381,544 75.6%

Total 8,110 100.0% 44,708,964 100.0% 9,054,928 100.0% 35,654,036 100.0% 504,587 100.0%Source: Own elaboration.

Communes Total Population Rural Population Urban Population SuperfaceTypology

Table 2. LAU2 Typology from the Urban/Rural grid cells: Communes, population distribution within the commune and surface.

12

Commune Urban/Rural typology

Page 13: Urban/Rural Areas

13

Administrative cities… Because LAU2 administrative boundaries are too

restrictive to identify directly urban communes with the wider concept of a city, we also define administrative cities as contiguous urban communes.

Page 14: Urban/Rural Areas

14

Administrative cities… As a result: 105 high density clusters (50.0% of

population) determine 220 urban or high density populated communes (53.7% of population), which in turn are mapped into 70 administrative cities.

This is similar (but not identical!), to the current definition used by the DG-Regio and the OECD.

We don’t have data (yet!) on commuting, so this is not taken into account.

Cities will be important when introducing accessibility.

Page 15: Urban/Rural Areas

15

Some Administrative Cities…

MadridPolycentric

city

ValenciaMonocentric

city

Barcelona: Polycentric city

Page 16: Urban/Rural Areas

16

Urban/Rural Typology from a 1-km2

population grid. This urban/rural typology from a 1-km2 is currently in

used by Eurostat/DG-Regio, and it is being implemented in EU-SILC and LFS.

In fact it has already been implemented in Spain by INE in the Household Budget Survey, so we can look at some socio-economic indicators:

Main drawback: This typology takes into account only the demographic factor, population density, but it is free from commune size administrative boundaries.

Urban Intermediate RuralIncome Per capita 8,821 € 9,594 € 8,526 € 7,562 €Gini Index (Inequality) 33.2% 33.1% 32.5% 32.4%Poverty Index (Head count) 20.0% 17.1% 20.1% 25.7%Source: Own elaboration from Spanish HBS 2011

Table 3. Economic indicators by Rural/Urban Typology based on population clustersTypologySpain

Page 17: Urban/Rural Areas

17

Extending the Urban/Rural Typology taking into account other factors… Population pressure is only one aspect of the urban/rural

landscape: It is a proxy for rural/urban influence.

However, urban/rural typologies should take into account the degree of human intervention of the landscape, that can be proxied by the share of artificial surfaces/build-up areas in land cover.

And this can be measured using land cover information without any reference to population.

Define: “open space” as agricultural, forest, natural areas, wetlands and inland water, and “close space” as built-up areas: artificial surfaces, including reservoirs.

Page 18: Urban/Rural Areas

18

Open/Closed space LAU2´s…

The communes are classified according to their share of artificial surfaces using the following rules:

1. Open space commune if at least 90% of its surface is “open space”. At the most its artificial surface account for 10% of its total surface.

2. Intermediate commune if its surface of “open space” is at least 75%, but lower than 90%. Its artificial surface is more than 10%, but at the most is 25%.

3. Close space commune if more than 25% of its surface is “closed space”, this is artificial surface. Less than 75% of its surface is “open space”.

Page 19: Urban/Rural Areas

19

SIOSE: The land cover data set. We use a national high resolution land cover data set:

SIOSE: Minimum Mapping Unit for artificial areas of 1 ha.

The process is very simple:1. Overlay the land cover data set with LAU2 boundaries.

2. Reclassify the land cover classes into “open” and “closed” spaces.

3. Recalculate surfaces, and determine the corresponding shares.

Page 20: Urban/Rural Areas

20

Open/Closed space LAU2´s…

Typology Number % Inhabitants % Km² %Closed space 351 4.3% 19,857,972 44.4% 8,425 1.7%

Intermediate space 779 9.6% 11,069,746 24.8% 34,243 6.8%Open space 6,980 86.1% 13,781,246 30.8% 461,920 91.5%

Total general 8,110 100.0% 44,708,964 100.0% 504,587 100.0%Source: Own elaboration.

Communes Population SurfaceTable 4 Commune typology using land cover information.

Page 21: Urban/Rural Areas

21

Crossing the typologies...

Number % Number % Number % Number %Urban 148 1.8% 55 0.7% 17 0.2% 220 2.7%

Intermediate 171 2.1% 376 4.6% 478 5.9% 1,025 12.6%Rural 32 0.4% 348 4.3% 6,485 80.0% 6,865 84.6%Total 351 4.3% 779 9.6% 6,980 86.1% 8,110 100.0%

Land CoverUrban Clusters (Demography) Closed space Intermediate space Open space Total

Table 5 Commune typology from Urban Clusters and Land Cover

Page 22: Urban/Rural Areas

22

We find that 80% of communes are Rural/Open, against less than 2% Urban/Closed, but in terms of population the importance is reversed: 39% of population live in Urban/Closed against 14% in Rural/Open communes.

As expected, we have more diversity in intermediate communes from both perspectives.

Crossing the typologies...

Inhabitants % Inhabitants % Inhabitants % Inhabitants %Urban 17,507,765 39.2% 5,007,128 11.2% 1,487,685 3.3% 24,002,578 53.7%

Intermediate 2,285,234 5.1% 5,252,111 11.7% 6,098,069 13.6% 13,635,414 30.5%Rural 64,973 0.1% 810,507 1.8% 6,195,492 13.9% 7,070,972 15.8%Total 19,857,972 44.4% 11,069,746 24.8% 13,781,246 30.8% 44,708,964 100.0%

Source: Own elaboration.

Table 6 Population in commune typology from Urban Clusters and Land Cover

Urban Clusters (Demography)

Land CoverClosed space Intermediate space Open space Total

Page 23: Urban/Rural Areas

23

We can see that 94% of Rural communes are classified as “Open space”, which indicates a high relationship between demographic density and built-up areas (correlation coefficient 0.73).

Population density versus built-up areas…

This relationship is far from linear. Non-linearity is clear for Intermediate and Urban communes.

The natural conclusion is that “population pressure” and “human intervention” play a different role in urban/rural relationships, which is specially true in “non-Rural”/”non-Open space” communes.

Page 24: Urban/Rural Areas

24

Accessibility as a differential factor among Rural Areas. As a discriminating factor among rural areas we study

its accessibility to cities.

From the set of accessibility indicators we choose (for the time being) “travel time”, but are in the process of extending the indicators, the transport modes and the type of analysis.

Main drawback: GIS network data is scarce and of very bad quality, specially the data sets coming directly from the National Geographical Institute (IGN).

We were forced to restrict the calculations to the road network, and a lot of time was devoted to create the “Network Dataset” (ND) at the required level of detail.

Page 25: Urban/Rural Areas

25

The “Network Dataset”… We experimented with 3 different data sets:

1. Official topographic maps (IGN, vector format): BCN200, scale 1:200.000.Advantages: Official data. Disadvantages: scale, broken network at many point and incomplete data.

2. Open Street Maps (OSM, vector format): Advantages: ArcGIS editor.Continuous updating, information increasing rapidly.Disadvantages: not official data, reference date unknown, quality uneven.

3. Google´s API´s (not GIS data):Advantages: Numeric data, continuous updating. Disadvantages: not official data, reference date unknown, quality uneven, queries limited by Google.

Page 26: Urban/Rural Areas

26

Eventually we used OSM, even if we were reluctant to used non-official data.

Conclusion: Open source free geographical data poses a lot of pressure to official mapping agencies.

Travel speed limits are defined for:1. Motorway: 120 Km/h.2. Trunk road: 110 Km/h.3. Primary road: 90 Km/h.4. Secondary road: 70 Km/h.5. Slip road: 70 Km/h.6. Local, tertiary, residential & living streets: 50 Km/h.

These speed limits are adjusted by slope and congestion in cities.

The “Network Dataset”…

Page 27: Urban/Rural Areas

27

Travel time (in minutes) is calculated as:

Shape_length in meters and Speed_limit in Km/h.

Travel time is affected by relief: the slope gradient on each road segment was derived from a Digital Elevation Model (DEM): NASA Shuttle Radar Topographic Mission (SRTM).

Also, because our dataset is quite detailed, tunnel segments were also taken into account, to avoid outliers in the slope index. In these cases, the Slope_index was substituted by a Tunnel_index in the previous formula.

The “Network Dataset”…

_ * _ * __

1000_

60

Shape length Slope index Congestion indexTravel Time

Speed limit*

Page 28: Urban/Rural Areas

28

Travel time (in minutes) is calculated as:

Shape_length in meters and Speed_limit in Km/h.

Congestion also affects travel time. The Congestion_index is defined when roads overlay with Urban Morphological Zones (UMZ), defined as “A set of urban areas laying less than 200 m apart”.

Those UMZ were obtained by us in a previous study from SIOSE land cover data set.

The “Network Dataset”…

_ * _ * __

1000_

60

Shape length Slope index Congestion indexTravel Time

Speed limit*

Page 29: Urban/Rural Areas

29

The “OD travel time matrix”… Given the ND data set, the next step is to construct the

Origin-Destination (OD) matrix by defining origins and destinations.

Origin: LAU2 (8,110). Population weighted (with the 1 km2 grid) centroids of local units.

Population weighting matters a lot in rural areas in order to locate people of local units at a single point. Roads are frequently quite far from geometric centroids, but they always cross populated cells.

UnweightedPopulation Weighted

Page 30: Urban/Rural Areas

30

The “OD travel time matrix”… Destinations: Urban Centers (HDC, 105). They are the

core in the determination of densely populated communes (220), which in turn aggregate to cities (70).Population weighted (with the 1 km2 grid) centroidsof HDC, even weighting is not now very important.

Hence, we eventually have a 8,110105 OD travel time matrix with the minimum travel time to reach the

nearest Urban Center, or if you prefer the nearest city.

Page 31: Urban/Rural Areas

31

Remote/Close to a city classification. Using as threshold 60 minutes, we classify a commune as

remote if the travel time to reach an Urban Centre, a City, is of at least 60 minutes.

Results are however quite sensitive to the threshold value, so further experimentation is probably necessary.

60 minutesthreshold Number %Close 6,185 76.3%Remote 1,925 23.7%Total 8,110 100.0%

45 minutesthreshold Number %Close 4,753 58.6%Remote 3,357 41.4%Total 8,110 100.0%

30 minutesthreshold Number %

Close 2,943 36.3%Remote 5,167 63.7%

Total general 8,110 100.0%Source: Own elaboration.

Sensitivity

Communes

Communes

Communes

Page 32: Urban/Rural Areas

32

Remote versus Close Rural Communes

Number % Number % Number % Number %Urban 148 1.8% 55 0.7% 17 0.2% 220 2.7%Intermediate 171 2.1% 376 4.6% 478 5.9% 1,025 12.6%Rural 32 0.4% 348 4.3% 6,485 80.0% 6,865 84.6%

Close 28 0.3% 307 3.8% 4,698 57.9% 5,033 62.1%Remote 4 0.0% 41 0.5% 1,787 22.0% 1,832 22.6%

Total 351 4.3% 779 9.6% 6,980 86.1% 8,110 100.0%Source: Own elaboration.

TotalLand Cover

Table 9 Close versus Remote Rural Communes

Urban Clusters (Demography) Closed space Intermediate space Open space

Using this threshold, 27% of rural communes are classified as remote.

This accounts for 24% of rural population, 3% of the total population.

Average size of rural/remote communes is just 700 inhabitants,

Page 33: Urban/Rural Areas

33

Next step… Next step is measuring economic performance by

commune typology, since for the moment we only have demographic information at this level of geographical detail.

For example, mean age for Spain is 42 years, mean age for rural communes is 49 years, and for rural/remote communes reaches 52 years.

After Census 2011 data is released we expect to apply Small Area Estimation (SAE) techniques to disaggregate economic variables and measuring economic performance for different typologies.

Page 34: Urban/Rural Areas

Many thanks for your attention.

34


Recommended